How to Find Your First Customers | Winning Your First 10 with Six Approaches

Hello, this is Ryuta Hamamoto from TIMEWELL.

Whenever I advise people on new businesses, there is one spot where the words always get stuck. "I get it, talking to customers is what matters." Up to here, everyone looks bright and confident. Then their voice drops a little and they add: "...so, these customers - where exactly are they?"

This is the wall of the first 10. You have the business idea, the hypothesis about the problem, and even a prototype in hand. And yet there is not a single person to show it to. When this state drags on for a few weeks, most people convince themselves the idea must be bad. But the cause is usually not the idea. It is simply the mistaken belief that first customers are something you "find." You will not find them by searching. Your first customers are people the founder wins over and brings in, one at a time. In this article I will break down how to do that persuading into six routes, in a form you can act on starting tomorrow.

If you want an objective look at where you are stuck in the business-building process before reading on, spend about 10 minutes taking stock with the AI Literacy Check first. It will make it easier to see which parts of this article to use.

The heaviest question that comes right after "go talk to customers"

The places where new businesses stall are not evenly distributed. In my experience, about 30 percent of the people who drop out do so at this one point: they cannot find their first customer. Not before the product is done, not after - they stop right before going to meet that first person. There are three main reasons.

The first is the fear of rejection. You reach out to a stranger and get ignored, or get a cold reply. To avoid that sting, people put it off with "let me prepare a little more first." That is running away in the name of preparation. The second is that there are too many methods to choose from. Cold email, social media, referrals, ads, events. Information is endless, so time slips by while you fail to decide where to start. The third is the most damaging. In a panic, you try everything at once and everything ends up half-done. This week 10 cold emails, next week a few social posts, then a couple of messages to acquaintances. None of them gets enough volume, so all you are left with is the false conclusion that "none of it worked."

Let me share one iron rule of the seed stage. Deliberately do things by hand that do not scale. This is a mindset almost every founder passes through in the earliest days¹. A system where you run ads and customers gather automatically is genuinely beautiful. But if you do that before you have found a repeatable way to sell, the cost of acquiring a single customer easily exceeds what that customer will ever pay you over their lifetime. Once acquisition cost tops customer lifetime value, the more you sell, the deeper the losses. The only way to avoid that collapse is, at first, for the founder to bring customers in one by one, by hand and on foot. It is gritty, exhausting, and unscalable. But there is something you can only get here: what customers are truly struggling with, and which words actually land when you explain your product. That resolution does not come from statistics on 100 people; it comes only from raw, direct exchanges with your first 10.

The six routes that reach customers - no single right answer, but there is a best one

The routes to your first 10 can be sorted into six. Think of it as six trails up the same mountain. Any of them gets you to the summit. But depending on your stamina, your gear, and the weather, the easiest climb changes. First, the big picture. The response rates are rough guides from experience and will vary by business and by counterpart. Grab the order of magnitude rather than the exact numbers.

Route	Best suited for	Rough response rate
Cold email / DMs	B2B companies, clear decision-makers	1-3% first reply (5-15% for those who customize)
Social media outreach (X, Instagram)	B2C, products that spread through empathy	Varies widely with post quality and consistency
Surveys and messaging	Regional SMEs, wary audiences	Fewer drop-offs the more gradual the steps
Referrals from people you know	You have a network, trust-driven products	5-10x cold outreach
Joining communities	A place where a specific interest group gathers	Trust builds in about three weeks with give-first
Going on-site (part-time work, volunteering)	You don't know the industry from the inside	Small volume, but the highest resolution

The point is that none of these six is inherently superior. Someone with zero network who suddenly aims for referrals has no one to refer them. Someone who dislikes writing and relies solely on cold email will never get the pen moving. The right answer is to pick the route that gets you in front of your first person fastest with the cards you already hold.

The three text-based routes (cold, social, surveys)

Cold email and DMs are the most repeatable route for B2B outbound. As long as you can build a list, you can stack up actions in proportion to the time you invest. The flip side: blasting a template as-is destroys both opens and replies. What works here is four principles. Keep it short (body under about 150 words). Include exactly one sentence with the recipient's specific name or company. Narrow your ask to one thing. Lead with what is in it for them. Spend five minutes rewriting just the placeholder parts of your template for one person. That five minutes triples your reply rate. Sending a template as-is is instant death.

Social media outreach shines for B2C and consumer products, especially those where people move on empathy. If your main customers are women - parenting or beauty, say - Instagram reaches them better than X. What to remember here is the power of a DM that signals you are in the same boat. Message people not as "a vendor trying to sell something" but as "someone dealing with the same struggle," and even a first Instagram DM can get a 20-30 percent reply rate. Conversely, if you use social outreach to reach a corporate decision-maker, they simply are not seeing that post. When the route and the counterpart's daily orbit are misaligned, all your effort swings and misses.

Surveys and messaging is the route for closing in gradually on wary counterparts. It works for regional SMEs and for audiences who hate being pitched out of the blue. Start with a light survey - "just one question, please" - then send a thank-you via messaging to those who answer, and slowly close the distance from there. Because you are not leaping straight to a sales meeting, you get to talk only after trust has accumulated.

The three relationship-based routes (referrals, communities, going on-site)

Referrals are the highest-ROI route at the seed stage. "5 to 10 times the response of cold outreach" is no exaggeration - a third party's trust rides along directly. But many people fail at asking for referrals. The cause is "dumping the work on the referrer." Just saying "could you introduce me to someone?" leaves them unsure what to say or how, so they cannot act. The key is to cut the referrer's effort to the absolute minimum. Along with your request, hand them a ready-to-forward message. All they have to do is copy and forward it. Adding one line - "if it doesn't feel right, please feel free to decline" - instantly lightens the psychological load on the person you are asking.

Joining communities means entering a place where people with a specific interest gather, banking trust, and only then talking business. The forbidden move here is starting promotion or DM sales the moment you join. You get branded as "the salesperson" in one shot and lose all credibility on the spot. For the first three weeks, commit entirely to giving. Answer questions, share information, help people who are stuck. Selling can wait; there is plenty of time for it once trust has built up.

Going on-site is the grittiest route and the one that raises your resolution the most. If you do not know the customer's industry well, get inside it through part-time work or volunteering. It works best precisely in worlds that are wary of outsiders, like long-established firms or traditional crafts. For counterparts like these, a handwritten letter reaches far better than a gift card or a web ad. If your target is students, going through a professor is a faster path than approaching them directly. Where does your counterpart usually spend their time, and by what customs do they come to trust someone? Observing that and choosing your route accordingly is the essence of using the six routes well.

Which route should be your first priority? (a situation-based cheat sheet)

By now you are probably thinking, "So which one do I choose?" To speed up the decision, here is a cheat sheet by customer type. Circle exactly one first priority. The second priority is something you add only after the first one is on track. You do not start them at the same time.

Customer type	First priority	Second priority	Support
B2B companies	Cold email / DMs	Referrals	Communities
B2C general	Social media outreach	Communities	Referrals
Women (parenting, beauty)	Instagram outreach	Same-boat DMs	Surveys and messaging
Regional SMEs	Surveys and messaging	Referrals	Going on-site
Long-established firms, crafts	Handwritten letters	Referrals	Going on-site
Students	Referral via a professor	Social media outreach	Communities

There is one trick to using this table. Pull out a sheet of paper and, when you hear "your first 10," write down the people whose faces come to mind. If it is zero, the text-based routes (cold email, social media) are your shortcut. If three or more come to mind, referrals and communities built on those relationships are faster. You are working backward from the assets you already have.

The other thing I want to stress: a route that does not fit the customs of your target's world backfires. Reach a corporate decision-maker on Instagram and they are not looking; send a digital gift to a long-established regional firm and you come across as lightweight. The circuits through which people extend trust differ completely by industry and generation. Choose the route your counterpart finds easy to receive, not the one you find easy to run. Get this wrong and results simply will not come, no matter how much effort you pour in.

Don't do everything. Pick one and run 100 actions

The biggest mistake in the first three months is running all six routes at once. Go parallel and the number of actions per route drops, none of them reaches the volume you need to improve, and you end up quitting all of them. So the principle is simple. Pick one and complete 100 actions over three months. For cold email, that is 100 messages; for social media, 12 straight weeks of posting; for referrals, 20 requests. Only when you hit that volume do you have the material to decide how to improve your subject lines and openers.

Why 100? So many people conclude "it doesn't work" after 10 emails, but that is far too weak as data. At a 1-3 percent reply rate, 10 emails have an expected value of 0.1 to 0.3 replies. Getting zero is entirely normal. With 100 emails you can expect 1 to 3 replies and one meeting from those. Only with that volume can you compare "does subject line A or B get opened more?" Giving up before you finish is like calling a dish bad without taking a single bite.

To make this concrete, let me walk through 90 days of a made-up business. Call it "Genba Note." The premise: an AI app for electrical and equipment contractors with 10 to 50 employees that automatically generates work reports and invoices from on-site photos and memos. The founder previously worked at an equipment manufacturer and knows five people at contracting firms. Running the cheat sheet, the customer is a B2B company that skews toward regional SMEs. First priority is cold email, second is referrals, support is communities. Because "five faces come to mind," the founder anchors on two lines to start - referrals and cold email. Social outreach won't reach a contractor's decision-maker, so it is off the table for now.

Month one is for raising resolution through referrals and going on-site. The founder sends the ready-to-forward referral request to those five contacts and books meetings with the directors or site managers at three firms. Tagging along at one company's site for half a day, the founder learns that the real pain is not the "hassle of writing reports" they had assumed, but the "cash-flow delay caused by the lag between the report and the invoice." That discovery would have been impossible without going on-site. Here they get a beta version tried for free and collect weekly feedback. The target is three early users. Month two, they complete 100 cold emails. They build a list of 30 firms on three axes - "electrical and equipment contracting, 10-50 employees, recent signs of hiring or a new location" - prepare two subject-line variants, and split-send. Body under 150 words, results stated concretely (a report going from 30 minutes to 3 per entry), two proposed time slots. Baking in a 1-3 percent reply rate, they keep sending, steadily. The target is seven firms cumulatively. Month three, they ask each onboarded firm to introduce "two peers with the same problem" and keep the funnel turning. "We'll discuss it internally" gets logged as a soft no and rolled into a three-month-later follow-up. The target is 12 firms cumulatively, and this is where they judge whether things are on track.

There are two lessons from these 90 days. Because they did not run all six routes and instead narrowed to two - referrals and cold email - they accumulated enough data (100 emails) to improve their cold-email subject lines. And going on-site, even once, has the power to rewrite a pain hypothesis from the ground up. Narrowing volume is not cutting corners. It means going deep enough on one thing to actually learn from it.

The real work starts once you get a response (Founder Sales and the polite no)

Once a reply comes in, the real work begins. Seed-stage sales meetings are run by the founder, not a salesperson. Since you cannot yet put your repeatable way of selling into words, the goal is to unearth that formula from customer reactions. For a 30-minute call, an easy-to-run split is 5 minutes of discovery (hearing their situation and struggles), 15 minutes of demo or proposed solution, and a final 5 minutes to lock in the next step. Rather than endlessly explaining your own product, spend most of the time drawing out the other person's words.

Here is a pitfall specific to Japan: the culture of not saying no to your face. "We'll discuss it internally" is close to a no about 80 percent of the time, and "it looks great, just not right now" leans even harder toward no. Mistake this politeness for a likely close and your pipeline fills with phantom prospects. The countermeasure is to rate each prospect's warmth on a five-point scale in every meeting. Hot is 5 (wants to sign right away), Cold is 1 (a clear no). Record vague replies as vague, and never spin them positively on your own. And the clearest yardstick of real intent is whether you can set the next meeting's date and time on the spot. "I'll email you some dates later" is usually a sign of a lost deal. Send the calendar invite during the call and lock it in.

Not letting things go silent matters too. Eighty percent of replies come not from the first message but from follow-ups. For those who haven't responded, schedule follow-ups at not-too-pushy intervals - 3 days, 8 days, 15 days, then 30 days later. This quarter's no may be next quarter's yes. Track everyone long-term, including those who declined, in a Google Sheet with about five columns (contact, situation, warmth, last touch, next action). And with people you have built a good relationship with, always ask for two introductions at the end of the conversation. Making this one move a habit keeps the top of your funnel from drying up.

Turning this kind of customer-acquisition design from a founder's individual skill into a repeatable team system is exactly the work we do together in WARP, our AI consulting service. Which route to focus on, which steps to hand off to AI, how to spot a polite no. We work alongside you on the very decisions of building a business. Until annual recurring revenue reaches something like the tens of millions of yen, the founder sells personally and documents the repeatable way to sell before handing it off. Withdrawing too early - the founder stepping away from sales - is the most wasteful failure of all.

Make AI your customer-acquisition partner (copy-and-paste prompts)

Here is the part I most want you to take away. Every step of customer acquisition speeds up dramatically when you use AI as a sparring partner. The prompts below are written so you can paste them straight into Claude or ChatGPT. Replace the bracketed parts with your own business.

First, a sparring session to decide which route to focus on.

You are a mentor who has supported customer acquisition for more than 200 new
businesses. I am about to find my "first 10 customers."
Based on my situation below, from the six approaches (cold email/DMs, social
media outreach, surveys and messaging, referrals, joining communities, going
on-site), recommend a first priority, a second priority, and a support route.
For each, give one line on "why it fits my situation," and also tell me which
route I should NOT do right now, with the reason.
Finally, give me three concrete ways to write down my "first 10" candidates on
paper today.

[Business] ___
[Customer type] one of: B2B companies / B2C general / women / regional SMEs /
long-established firms and crafts / students
[Current network] ___ people I know who are close to the target
[Hours available per week] ___ hours
[Budget for incentives] ___ yen per month
[Strengths/weaknesses in writing vs. face-to-face] ___

Once you have settled on cold email, next you produce the full email set at once.

You are a sales pro who gets 10-15% reply rates on B2B cold email. For my
business below, produce the following four things.
(1) The angles and example search keywords for building a list of 30 companies
on three axes: "industry x role x recent trigger (procurement, hiring, new
service, etc.)"
(2) Five subject lines of 40 characters or fewer (mix question form, naming the
recipient, and mentioning a trigger)
(3) One email body of 100 words or fewer (include one sentence with the
recipient's specific name / exactly one ask / lead with what's in it for them /
propose two time slots)
(4) A three-message follow-up sequence for non-responders at 3, 8, and 15 days
Salesy tone, template smell, and self-centered writing are all forbidden; use
the recipient's name/company twice as often as your own name/company.

[Our category / service] ___
[Concrete results (numbers) we can deliver] ___
[Similar customers / track record] ___
[Assumed pain of the recipient] ___

When responses are poor after you start sending, have AI diagnose where things are stuck. The key here is to narrow the fix to one place.

I have sent ___ cold emails. My results are [open rate ___% / reply rate ___%
/ positive replies ___% / meetings booked ___ / deals won ___].
(1) Diagnose each metric as "good / needs work / danger" (rule of thumb: open
rate above 40% is good and below 25% is danger; reply rate above 10% is good
and below 5% is danger).
(2) Identify exactly one bottleneck and decide the order in which to fix
"subject line -> opener personalization -> CTA -> target list" (do not change
several at once).
(3) Give two A/B tests I can run over the next 30 emails, each changing exactly
one thing, concretely.
(4) Check whether I am misreading soft nos like "we'll discuss it internally"
as a yes - I'll paste my reply text for you to check.

[Subject lines / bodies I sent] ___
[Replies I received (if any)] ___

When you ask for a referral, have AI build a forward-ready set that gets the referrer's effort close to zero.

I've heard referrals are 5-10x as effective as cold outreach at the seed stage.
Based on the following, create a forward-ready set that cuts the referrer's
effort to the absolute minimum.
(A) The request message to send the referrer (consideration for the abruptness
of the ask + one line on why this person + an explicit note that they may
decline if it doesn't feel right)
(B) A body the referrer can copy-paste and forward directly to the target (a
one-line intro of me + a reason directly tied to the target's work + a request
for just 15-30 minutes + an instruction to CC my email)
Finally, add ideas for thanking the person who referred me, plus a template for
reporting back the outcome after the introduction.

[Referrer (who I'm asking)] ___
[Target of the referral, their affiliation and role] ___
[One-line description of my business] ___
[Reason directly tied to the target] ___

Before a meeting, role-playing with AI as the prospect keeps you from freezing up in the real thing.

You are playing my prospective customer. I am a seed-stage founder practicing a
30-minute Founder Sales call.
The flow is "5 minutes discovery -> 15 minutes of Q&A around a demo -> 5 minutes
to lock in the next step."
Behave as the following person, who is busy and skeptical but genuinely has the
problem, and naturally weave in Japan-specific soft rejections ("we'll discuss
it internally," "maybe in next year's budget," etc.).
For each thing I say, give me (1) your reply as the customer, (2) a behind-the-
scenes warmth rating (Hot 5 to Cold 1) with the reason, and (3) a note on
whether my question or demo landed and whether I got self-centered - every turn.
Finally, judge whether I managed to lock in the next meeting's date and time on
the spot, and if not, give me a model example of how to rephrase.

[Customer role: industry, title, company size] ___
[Assumed problem the customer has] ___
[My service] ___

The point I want to nail down here: AI is not a substitute for the customer. The role-play partner and the reply diagnosis are only for preparation and organization. Only the raw, real reactions of actual customers can teach you your repeatable way to sell. Think of AI as a partner that speeds up your preparation so you can step onto that real stage again and again. If you want to make this kind of AI fluency a team standard rather than a matter of individual talent, start by measuring where you stand with the AI Literacy Check.

Summary

We have traced how to find your first customers - the six routes and how to run them. Here are the key points, each paired with a step you can take tomorrow.

• First customers are not something you find; they are people the founder wins over one by one. Treat it as a period of deliberately doing unscalable work by hand.
• There are six routes and no single right answer. From your customer type and your own cards, choose exactly one first priority.
• Don't do everything. Pick one and run 100 actions over three months. Only once you have enough volume to improve can you fix your subject lines and openers.
• When responses come, rate warmth on a five-point scale every time and don't misread a polite no as a yes. Lock in the next meeting's date and time on the spot.
• AI is your partner for route selection, template generation, funnel diagnosis, referral messages, and call practice. But it is not a substitute for the customer.

One last thing. Until you can meet that first person, debating whether the business idea is good or bad is meaningless. You simply do not yet have enough reality in hand to judge. So today, circle one of the six routes and decide on one of these: send one email tomorrow, ask one person, or make one post. By the time you have met your tenth customer, you will be able to say half of what is written in this article in your own words.

If you want to re-map the entry point of building a business, head to The Complete Guide to Building a New Business. If you are stumbling just before unearthing the customer's true problem, Finding Customer Problems and the Empathy Map is a solid foundation. For narrowing down who you deliver to, see the Value Proposition Canvas, and for estimating market size, How to Size a Market.

If you want to think out loud about which route to focus on, or tailor cold-email subject lines and referral requests to your specific business, come talk with us through a WARP consultation.

---

References

Footnotes

1. The source for the idea of doing things that don't scale. Paul Graham, "Do Things That Don't Scale" (2013, essay). It argues for the importance of the founder acquiring customers one by one, by hand, in the earliest days. ↩